Perceptions and Attitudes of Emergency Physicians in Saudi Arabia Regarding Medical Use of Artificial Intelligence.

BACKGROUND: Artificial intelligence (AI) is increasingly recognized for its potential to support clinical decision-making and workflow optimization in emergency medicine (EM). However, the perspectives of emergency physicians, particularly in Saudi Arabia, regarding its integration remain understudied. OBJECTIVE: To assess the attitudes, familiarity, perceived advantages, and concerns of Saudi emergency physicians regarding the application of AI in clinical emergency practice. METHODS: A descriptive cross-sectional study was conducted between February and April 2025 using a self-administered online questionnaire distributed to emergency physicians across Saudi Arabia. Inclusion criteria were active clinical engagement and at least one year of EM experience. Descriptive and inferential statistics were performed using SPSS version 26. RESULTS: Among 101 participants, the majority were male, 71 (70.3%), and aged 25-29 years, 71 (70.3%). A total of 51 (50.5%) agreed that they were familiar with AI, while 97 (96%) believed that AI has useful applications in medicine. However, 41 (40.6%) disagreed that AI has superior diagnostic ability compared to human physicians, while 35 (34.7%) strongly disagreed that AI could replace their roles. The most cited advantages of AI were access to high-quality real-time clinical data, 75 (74.3%), absence of emotional or physical exhaustion, 69 (68.3%), and faster healthcare processes, 64 (63.4%). In cases of disagreement between physician and AI, 95 (94.1%) preferred the physician's opinion. Key concerns included AI's unreliability in unpredictable situations, 72 (71.3%), lack of flexibility in application, 64 (63.4%), and inability to handle controversial topics, 69 (68.3%). Over half, 54 (53.47%), expressed positive attitudes toward AI in EM. Attitudes were significantly more positive among physicians from the northern region (β=2.8, 95% CI: 0.37-5.1, p=0.024) and western region (β=2.0, 95% CI: 0.30-3.7, p=0.022) compared to the central region. CONCLUSION: Saudi emergency physicians generally express positive attitudes toward AI, particularly in supporting clinical tasks, but remain cautious about its diagnostic authority and legal implications. Regional variation suggests the need for tailored educational and policy strategies to promote AI adoption in emergency settings.